Digital tissue segmentation and viewing
Methods and systems for representing a tissue segmentation from a source digital image computationally generate, from a source digital image of an anatomic region, a digital tissue segmentation visually indicating regions of interest corresponding to an abnormal condition associated with at least portions of the anatomic region. The source image and the tissue segmentation may be alternately displayed in registration on a mobile device at a gesturally selected magnification level.
1. A method of computationally representing and displaying a digital tissue segmentation image and a source digital image, the method comprising the steps of:
computationally generating, from a source digital image of an anatomic region, a digital tissue segmentation image visually indicating regions of interest corresponding to a physiologic condition associated with at least portions of the anatomic region;
alternately displaying each of the source digital image and the digital tissue segmentation image or corresponding portions thereof in registration at a first gesturally selected magnification level; and
responsive to a gestural selection of a second magnification level higher than the first magnification level, alternately displaying each of the source digital image and the digital tissue segmentation image or corresponding portions thereof in registration at the second gesturally selected magnification level.
2. The method of claim 1 , wherein the digital tissue segmentation image designates, and colorwise distinguishes, high-precision regions of interest and high-recall regions of interest superimposed on corresponding regions of the the source digital image.
3. The method of claim 1 , wherein the digital tissue segmentation image is an overlay.
4. The method of claim 1 , wherein the source digital image and the digital tissue segmentation image have substantially identical dimensions.
5. The method of claim 1 , further comprising computationally analyzing one or more of the regions of interest to identify the physiologic condition.
6. The method of claim 5 , further comprising computationally analyzing one or more of the regions of interest to identify a subtype associated therewith.
7. The method of claim 6 , wherein the subtype is a cancer subtype.
8. The method of claim 6 , wherein the subtype is a cancer severity level.
9. The method of claim 1 , wherein the digital tissue segmentation image is computationally generated at least in part by:
computationally generating a plurality of overlapping subimage regions of the source digital image;
computationally analyzing at least some of the subimage regions using a predictor to produce classification probabilities therefor;
computationally combining, for each pixel of the source image contained within a plurality of subimage regions, the classification probabilities corresponding to all of the analyzed subimage regions containing the pixel; and
computationally generating the digital tissue segmentation image based at least in part on the pixelwise combined predictions.
10. The method of claim 1 , wherein the digital tissue segmentation image is computationally generated at least in part using an object detector.
11. A system comprising:
a remote computer configured to store and transmit a digital tissue segmentation image computationally generated from a source digital image of an anatomic region, the digital tissue segmentation image visually indicating regions of interest corresponding to a physiologic condition associated with at least portions of the anatomic region; and
a local device configured to (i) receive the digital tissue segmentation image from the remote computer, (ii) alternately display each of the source digital image and the digital tissue segmentation image or corresponding portions thereof in registration at a first gesturally selected magnification level, (iii) receive a gestural selection of a second magnification level higher than the first magnification level, and (iv) alternately display each of the source digital image and the digital tissue segmentation image or corresponding portions thereof in registration at the second gesturally selected magnification level.
12. The system of claim 11 , wherein the digital tissue segmentation image designates, and colorwise distinguishes, high-precision regions of interest and high-recall regions of interest superimposed on corresponding regions of the the source digital image.
13. The system of claim 11 , wherein the digital tissue segmentation image is an overlay.
14. The system of claim 11 , wherein the source digital image and the digital tissue segmentation image have substantially identical dimensions.
15. The system of claim 11 , wherein the remote computer is configured to generate the digital tissue segmentation image and computationally analyze one or more of the regions of interest to identify the physiologic condition.
16. The system of claim 15 , wherein the remote computer is further configured to computationally analyze one or more of the regions of interest to identify a subtype associated therewith.
17. The system of claim 16 , wherein the subtype is a cancer subtype.
18. The system of claim 16 , wherein the subtype is a cancer severity level.
19. The system of claim 11 , wherein the remote computer is configured to computationally generate the digital tissue segmentation image at least in part by:
computationally generating a plurality of overlapping subimage regions of the source digital image;
computationally analyzing at least some of the subimage regions using a predictor to produce classification probabilities therefor;
computationally combining, for each pixel of the source digital image contained within a plurality of subimage regions, the classification probabilities corresponding to all of the analyzed subimage regions containing the pixel; and
computationally generating the digital tissue segmentation image based at least in part on the pixelwise combined predictions.
20. The system of claim 11 , wherein the remote computer is configured to computationally generate the digital tissue segmentation image at least in part using an object detector.